This article presents a theoretic framework, ''patient practice variat
ion,'' currently missing from the investigations of medical care varia
tion. Resource utilization, immediate, and long-term outcomes may be b
etter explained by including utilities under control of the patient in
to small area variation studies. This especially may be important in t
he area of prenatal and maternal care and certain chronic diseases, su
ch as diabetes. The quality of care measures must incorporate patient
satisfaction and quality of life in addition to more objective physiol
ogic outcomes.